Growth in reading after the point of decoding fluency is almost all in reading comprehension. There should be some headroom in the MAP 2-5 for at least a little growth in reading comprehension, but not as much for someone in the mid-90s %ile. The mean fall-to-winter reading growth for grade 3 was 7 points in the norm study, but less for already high-performing students.
To be fair, with regression to the mean, a child with 99%ile verbal cognition would be expected to have achievement in the low 90s, or even lower, so his achievement isn't discrepant. Stagnation continues to be the more important point.
As to his program, it may be that the exact form in which they expect reading comprehension to be demonstrated is not exactly that to which he is accustomed. I think we've discussed elsewhere on the forum the specific nature of Common Core reading comprehension tasks. Some of that may also be happening on the CC-aligned MAP testing.
